My last project involved building a rather ambitious pipeline to process huge numbers of twitter tweets, both in real time and cached. It was astonishing what two devs can achieve these days using off-the-shelf open-source tools. We ran the cluster of machines on Amazon's AWS, which made scaling the job-servers a fairly trivial affair.
After completing an MSc in Knowledge Based Systems at Sussex University I hung around to get a doctorate with the Centre for Computational Neuroscience and Robotics. My doctorate involved applying techniques from Evolutionary Robotics to the problem of insect navigation - in a nutshell why bees and wasps forage for food differently than ants. I stayed on the academic conveyor belt for a few post-docs, which involved, among other things, becoming the greatest bumblebee trainer alive and developing a novel robot control architecture using chemical reaction-diffusion systems.
I spent some time as co-founder and CTO building a couple of Internet companies:
Showmedo is a site where one can watch peer-produced videos to learn about programming and Open Source Software. It has the largest collection of Python programming videos on the web. Showmedo forced me to develop some web-dev chops as I was responsible for front and back end development.
I recently spent some time building Strongsteam, an ambitious web-based, RESTful API to make using algorithm's from Computer Vision, Natural Language Processing etc.. easy. This involved building a scalable architecture that could adapt to the heavy processing requirements. Python was the glue language, with fast C/C++/Fortran libraries doing the heavy lifting.